Video Face Recognition based on External General Dictionary
Qing Liu, Shaohu Peng, Jiadong Wang, Xiao Guang Hu · 2015
Face recognition with single sample per person (SSPP) was a very challenging task because SSPP cannot reflect the characteristics of the test samples (eg.illuminations, pose, expressions and disguises) so that it is difficult to predict the facial variations of a query sample by the gallery samples.Considering the fact that the intra-class facial variations can be shared across different subjects, a novel algorithm was put forward.This algorithm took generic variation dictionary to solve the problem of lack the number of sample.This method need create two different dictionary.Generic variation dictionary D was built by the different group of pictures and gallery dictionary X was built from the training samples.Using half-quadratic optimization function to solve optimization problem.The experimental results showed that the recognition rate of the algorithm was superior to recognition rate of some traditional methods.